Rapidly analyzing parametric resonance and manufacturing yield of MEMS 2D scanning mirrors using hybrid finite-element/behavioral modeling

Abstract

A new hybrid 3D finite-element/behavioral-modeling approach is presented that can be used to accurately predict the nonlinear dynamics (parametric resonance) in electrostatically driven 2D resonant MEMS scanning mirrors. We demonstrate new levels of accuracy and speed for thick SOI scanning mirrors with large scanning angles and validate the modeling approach against measurement on a previously fabricated scanning mirror. The modeling approach is fast and treats the design parameters as variables thus enabling rapid design iterations, automatic sensitivity and statistical yield analyses, and integration with system and circuit simulators for coupled MEMS-IC cosimulation.